Why this poison is interesting

Laburnum is the library's study in the gap between a poison's reputation and its behaviour. Laburnum anagyroides — the golden-chain tree — is routinely named among the most dangerous garden plants, its pea-like seed pods invite handling by children, and generations of gardening advice have treated it as lethal. Yet the great majority of childhood laburnum ingestions produce nothing worse than vomiting, and documented fatalities are very rare.³ The interesting question is why both things are true at once: a genuine toxin, and a usually mild course.

The answer is dose and pharmacology. The toxin, cytisine, is a nicotinic agonist — and it is the same molecule now licensed as a smoking-cessation medicine and the structural template from which varenicline was designed.¹ A molecule that can be given deliberately, at a controlled dose, as a therapeutic partial agonist is not one that kills reliably at the small, erratic doses a child obtains from chewing a few seeds. The poison is real; the usual exposure is small; and the mechanism accounts for the vomiting that limits how much toxin is retained.

The toxic principle

Cytisine binds nicotinic acetylcholine receptors as an agonist. At the receptor it behaves as a high-affinity partial agonist — which is precisely the property exploited therapeutically, where partial agonism blunts withdrawal while occupying the receptor.¹ In overdose the relevant behaviour is full nicotinic stimulation across autonomic ganglia, the neuromuscular junction and the CNS.

The early, prominent vomiting is itself partly protective: a toxin that reliably makes a child sick tends to be expelled before a large dose is absorbed, which is one reason the usual course is mild. It is also why the emetic phase, unpleasant as it is, is not the dangerous part; the dangerous part is the rare large retained ingestion that reaches the depolarising-block end of the curve.

Toxicokinetics

Because cytisine is a licensed medicine, its therapeutic pharmacokinetics are better described than most botanical toxins — it is orally absorbed, renally eliminated and short-acting at therapeutic doses.¹ The overdose kinetics are less formally studied, but the therapeutic profile frames what to expect.

Cytisine — a botanical toxin that is also a characterised drug
Parameter	Therapeutic	In overdose	Why it changes
Absorption	Orally absorbed; seed coats are hard, so an intact swallowed seed may release little toxin while a chewed one releases more¹	Chewing is what matters — the number of seeds swallowed whole overstates the dose absorbed	This is why ingestion history should ask whether seeds were chewed or swallowed whole, and why counts of intact seeds passed can be reassuring.
Distribution	Crosses the blood–brain barrier — the basis of both its central cessation effect and its central overdose features	Not separately quantified in overdose	CNS penetration is why large ingestions can produce seizures and reduced consciousness, not only autonomic features.
Metabolism / elimination	Predominantly renal excretion, largely unchanged; short-acting at therapeutic doses¹	Elimination is renal, so significant renal impairment would prolong effect	The short duration at therapeutic dose is consistent with the usually self-limiting course of a small ingestion.
Dialysability	—	No established role in management	The usual case needs observation and antiemesis; the rare severe case needs supportive management, not extracorporeal removal.

Metabolism and the metabolites

Cytisine is active as ingested — there is no bioactivation and therefore no latent phase — and it is eliminated largely unchanged in the urine, so the metabolism section is, as for the other directly-acting botanical toxins, short by nature.¹ The mechanistically important point is not a metabolite but the structure–activity link to modern medicine: cytisine's nicotinic partial agonism was the pharmacophore from which varenicline was engineered, so the plant toxin and the drug are two points on one line.

Cytisine — toxin and template

Cytisine (laburnum seed)Already active. Nicotinic partial agonist¹
Autonomic, neuromuscular and central nicotinic effectsDose-dependent — mild GI at small dose, nicotinic crisis at large
Cytisine in urineShort-acting at therapeutic dose¹

Elimination and accumulation

Laburnum poisoning is a single-exposure event with no accumulation story; the short, renally-cleared kinetics mean that a patient who has passed through the emetic phase without progressing to nicotinic-crisis features is generally declaring a mild ingestion. The corollary is the standard one for a nicotinic agonist: the window in which severe features could still appear is early, so the observation period is set by the clinical trajectory and NPIS advice rather than by any measured concentration.

Target organs — and why those

Gastrointestinal tract and autonomic system

TargetNicotinic receptors in autonomic ganglia and the chemoreceptor trigger zone

Why hereNicotinic stimulation of autonomic ganglia and central emetic pathways produces the nausea, vomiting, salivation, sweating and tachycardia that dominate the usual laburnum ingestion. This is the mild, self-limiting face of the poison, and the vomiting limits further absorption. Established

At the bedsideNausea and vomiting within an hour or two is the typical and often the only feature; supportive antiemesis and observation usually suffice.

Neuromuscular junction (large ingestions)

TargetNicotinic receptors at the motor endplate

Why hereAt a large enough dose, sustained nicotinic occupation produces a depolarising block, so the picture can turn from stimulation to weakness — the same end of the spectrum as poison hemlock and nicotine. This is the dose-dependent severe face, and it is uncommon. Inferred

At the bedsideWeakness and, at the extreme, respiratory compromise from neuromuscular block; only expected after substantial ingestion.

Central nervous system (large ingestions)

TargetCentral nicotinic receptors, reached because cytisine crosses the blood–brain barrier

Why hereCentral nicotinic stimulation can produce agitation, reduced consciousness and, in severe poisoning, seizures; these central features track dose and are the reason a large ingestion is treated more seriously than the seed count alone would suggest. Inferred

At the bedsideReduced consciousness or seizures indicate a significant ingestion and warrant full supportive management with NPIS input.

Timeline of effects

Laburnum ingestion — usually stops at the first phase

Time

What you seeWhat is happening

0–2 hEmetic phase
What you seeNausea, vomiting, salivation, sweating, tachycardia. For most childhood ingestions this is the whole illness.
What is happeningNicotinic stimulation of autonomic ganglia and central emetic pathways; the vomiting limits further absorption.
Hours (large ingestions only)Nicotinic crisis
What you seeWeakness, hypotension, bradycardia, reduced consciousness, seizures; respiratory compromise at the extreme.
What is happeningA large retained dose drives sustained receptor occupation towards depolarising block and central toxicity — the uncommon severe course.
RecoveryResolution
What you seeSymptoms settle as the short-acting toxin is cleared; most patients recover fully.
What is happeningRenal elimination of largely unchanged cytisine; no accumulation and no delayed organ injury.

What the mechanism predicts at the bedside

Most laburnum ingestions are mild, and the mechanism says why — early vomiting limits absorption, and the small erratic dose a child obtains is far below the therapeutic-then-toxic curve of a nicotinic agonist.¹ Mild does not mean harmless, but it does mean the usual case is observation and antiemesis.
Ask whether the seeds were chewed. Intact swallowed seeds release little toxin; the dose that matters comes from chewed seeds, so the history should distinguish them and a count of whole seeds passed can be reassuring.
Serious features are dose-dependent and point to a large ingestion. Weakness, seizures or reduced consciousness indicate a substantial dose and the depolarising-block/central end of the nicotinic spectrum, and warrant full supportive care with NPIS.²
It is the same molecule as a cessation drug, which frames the danger correctly — a controllable therapeutic partial agonist, not a reliably lethal poison, but one that can cause a genuine nicotinic crisis at high dose.¹
There is no antidote and no level. Management is supportive and trajectory-guided, escalating only if the rare severe course declares itself.

The antidote, from the poison's side

Laburnum has no specific antidote, and for the usual mild ingestion none is needed — the management is antiemesis, fluids and observation while a short-acting toxin clears. The mechanistic interest lies in what an antidote would have to do, and why none exists.

Critical appraisal

The 'usually mild' claim is well supported but not a licence for complacency. Large observational experience and the pharmacology agree that most childhood ingestions are mild; the rare severe case is real and dose-dependent, and the risk assessment is a TOXBASE decision, not a reason to dismiss the ingestion.³ Established
The inflated admission statistics are a documented artefact. The often-quoted large admission figure was a single-region extrapolation whose own author attributed many admissions to the plant's reputation rather than to illness — a caution about reading historical poisoning statistics literally.³
Severe-case features are inferred from nicotinic pharmacology and fatal reports. The depolarising-block and central features expected at high dose follow from nicotinic agonist pharmacology and are described in rare severe and fatal cases, rather than being characterised in a controlled series of laburnum specifically.² Inferred
The cytisine-as-medicine framing must not be over-read. That cytisine is a safe, effective cessation drug at controlled dose does not make a large plant ingestion safe; it explains the shape of the dose–response, no more.

References

1
Gotti C, Clementi F. Cytisine and cytisine derivatives. More than smoking cessation aids. Pharmacological Research 2021;170:105700. PMID 34087351.
2
Musshoff F, Madea B. Fatal cytisine intoxication and analysis of biological samples with LC-MS/MS. Forensic Science International 2009;186:e1–4. PMID 19217730.
3
Bramley A, Goulding R. Laburnum "poisoning". British Medical Journal (Clinical Research Edition) 1981;283(6301):1220–1. PMID 6797517. (The single-region admissions extrapolation and the usually-mild childhood course.)

Laburnum (cytisine)

At a glance